How to build a Kubeflow Pipeline

Experiment with pipeline samples→

Want to learn how to create an ML application from Kubeflow Pipelines? In this episode of Kubeflow 101, we show you how to build a Kubeflow Pipeline from the ML model we explored in the last episode. Moreover, we give you a walkthrough of how to create, test and deploy your ML application in Kubeflow Pipelines. Watch to learn how Kubeflow Pipelines can bring orchestration to complex workflows when working with ML applications.

Last episode →

0:00 – Intro
0:40 – Pipeline steps overview
1:09 – Import dependencies, define constants
1:45 – Download trainer data
2:10 – Train the model
3:04 – Deploy the model
3:22 – Define and submit the Kubeflow pipeline
3:56 – Compile and share the pipeline
4:26 – Conclusion

Watch more episodes of Kubeflow 101 →
Subscribe to the GCP Channel →

Product: BigQuery, Kubeflow, Kubeflow Pipelines; fullname: Stephanie Wong;


Duration: 00:04:59
Publisher: Google Cloud
You can watch this video also at the source.